Figure 292929 Diversity of cover species - figure details are explained in the caption of Figure 12. 2000 2002 2004 2006 2008 Manly Stanmore Waiwera 20 15 10 Campbells Torbay Long Bay 20 Average diversity per site per diversity Average 15 10 2000 2002 2004 2006 2008 2000 2002 2004 2006 2008 Year Figure 303030 Evenness of cover species –- figure details are explained in the caption of Figure 12. 2000 2002 2004 2006 2008 Manly Stanmore Waiwera 0.30 0.25 0.20 0.15 Campbells Torbay Long Bay 0.30 0.25 Average evenness (E[var]) per per site evennessAverage(E[var]) 0.20 0.15 2000 2002 2004 2006 2008 2000 2002 2004 2006 2008 Year Long Bay Marine Monitoring Programme: Report 1999 – 2008 45 Figure 313131 CCA cover–- figure details are explained in the caption of Figure 12. 2000 2002 2004 2006 2008 Manly Stanmore Waiwera 80 60 40 20 Campbells Torbay Long Bay 80 Average %cover per per site %cover Average 60 40 20 2000 2002 2004 2006 2008 2000 2002 2004 2006 2008 Year Figure 323232 Sediment cover - figure details are explained in the caption of Figure 12. 2000 2002 2004 2006 2008 Manly Stanmore Waiwera 80 60 40 20 0 Campbells Torbay Long Bay 80 Average %cover per per site %cover Average 60 40 20 0 2000 2002 2004 2006 2008 2000 2002 2004 2006 2008 Year TR 2009/019 Subtidal Reef Ecological Monitoring: 1999 - 2008 46 4.2.3.3 Temporal patterns in trapped sediment Sediment trap rate There was strong evidence of a difference in sediment trap rate between bays when averaged over all sampled time (p < 0.001). Tukey’s HSD pair-wise comparisons revealed a gradient in average trapped sediment rate, as follows: C > L > W > T > M > S (significant differences between bays are shown in Table 12 and Figure 33). All bays showed significant evidence ( p < 0.05) of a decline over time in sediment trap rate (Figure 34). Manly and Torbay have had a comparatively small but steady decline in trapped rate. In contrast, the other bays showed a non-linear change over time – a large decline before 2004, followed by a more modest recovery and relative stability in the last few years. There was evidence that ENSO was negatively correlated with sediment trap rate at Campbells, Torbay, and Waiwera ( p < 0.05), i.e. sediment trap rate was generally higher during La Niña years. Table 121212 Differences in sediment trap rate (ranked North to South) between bay averaged over time. Bays sharing a group letter are not significantly different, e.g. Campbells and Long Bay are not significantly different (they share group membership A); similarly Long Bay and Waiwera are not different (both members of B), but Campbells and Waiwera are different. Waiwera Stanmore Manly Long Bay Torbay Campbells GroupGroupssss:::: trap B C C A & B C A rate Long Bay Marine Monitoring Programme: Report 1999 – 2008 47 Figure 333333 Log (average sediment trap rate) for each bay. Letters indicate group membership (see Table 12 for a detailed explanation). Log( sediment trap rate) Waiwera Stanmore Manly Long Bay Torbay Campbells North South 0 -0.2 -1) -0.4 .day 2 -0.6 -0.8 -1 A B A & B log(gm . cm-log(gm . -1.2 C -1.4 C C -1.6 Figure 343434 Scatter plot showing the change in the log (sediment trap rate) over time (averaged for each bay). Manly Stanmore Waiwera 0 -1 -2 -3 )] -1 -4 .day 2 -5 Campbells Torbay Long Bay 0 -1 Average log(trap[g/(cm rate -2 -3 -4 -5 Jan00 Jan01 Jan02 Jan03 Jan04 Jan07 Jan08 Jan00 Jan02 Jan03 Jan04 Jan07 Jan00 Jan01 Jan02 Jan03 Jan04 Jan05 Jan06 Jan07 Jan08 Jan05 Jan06 Jan01 Jan05 Jan06 Jan08 TR 2009/019 Subtidal Reef Ecological Monitoring: 1999 - 2008 48 Fine sediment trap rate There was strong evidence of differences in the fine sediment trap rate (i.e. sediment less than 63 µm) when averaged over time amongst the bays ( p < 0.001). Tukey’s HSD pair-wise comparisons revealed a gradient in average trapped sediment rate, as follows: C > L > W > T > M > S (significant differences between bays are shown in Table 13 and Figure 35). The change in fine sediment trap rate declined between 2002 and 2005, but has been comparatively stable since that year (Figure 36). There is no evidence that ENSO was correlated with fine sediment trap rate at any of the LBMMP bays. Table 131313 Differences between bay fine sediment trap rate (<63 µm) (ranked North to South) averaged over time. Bays sharing a group letter are not significantly different (see Table 12 caption for an example). Waiwera Stanmore Manly Long Bay Torbay Campbells Groups: rate of B B B B B A fine sediment Figure 353535 Log (average trapped fine sediment rate) for each bay. Letters indicate group membership (see Table 12 for a detailed explanation). Average trapped fine sediment rate 0.06 A 0.05 -1 0.04 B .day B B 2 0.03 B B 0.02 gm . cm- gm . 0.01 0 Waiwera Stanmore Manly Long Bay Torbay Campbells North South TR 2009/019 Subtidal Reef Ecological Monitoring: 1999 - 2008 49 Figure 363636 Scatter plot showing changes over time in the trapped fine sediment rate (<63 µm) over time (averaged for each bay). Note: measurement of fine sediment started later in LBMMP, so there is a different time scale to Figure 34. Manly Stanmore Waiwera -2 -3 )] -1 -4 .day 2 -5 Campbells Torbay Long Bay -2 Average log(trap Average fines[g/(cm rate -3 -4 -5 Jan05 Jan06 Jan07 Jan08 Jan02 Jan03 Jan04 Jan05 Jan06 Jan08 Jan02 Jan03 Jan05 Jan06 Jan08 Jan02 Jan07 Jan04 Jan07 Jan03 Jan04 TR 2009/019 Subtidal Reef Ecological Monitoring: 1999 - 2008 50 Fine sediment proportions There was strong evidence of differences among the bays in the proportion of fine sediment (less than 63 µm) in traps, when averaged over time (p < 0.001). Tukey’s HSD pair-wise comparisons revealed a gradient in average trapped sediment rate, as follows: T > S > M > C > L > W (significant differences between bays are shown in Table 14 and Figure 37). All bays show the same general (non-linear) pattern in the proportion of trapped fine sediment over time: (i) a sharp drop between 2002 and 2004, followed by (ii) a slight increase until 2006-7, (iii) then a comparatively stable period in the last year for most bays (Waiwera showed another decline in fine sediment proportions) (Figure 38). ENSO was positively correlated with the proportion of fine sediment; that is, the proportion of trapped fine sediment was generally lower during La Niña years. Significant positive correlations ( p < 0.05) were found at Campbells, Long Bay, Manly and Waiwera. Table 141414 Differences between bays in the proportions of fine sediment (<63 µm) averaged over time (ranked North to South). Bays sharing a group letter are not significantly different (see Table 12 caption for an example). Waiwera Stanmore Manly Long Bay Torbay Campbells Groups: %%%fine%fine C A & B A, B & C C A B & C sediment Figure 373737 Average proportion of fine sediment) for each bay. Letters indicate group membership (see Table 12 for a detailed explanation). Average proportions of trapped fines (< 63 μm) 60% A & B A, B & C A 50% C C B & C 40% 30% Percent 20% 10% 0 Waiwera Stanmore Manly Long Bay Torbay Campbells North South TR 2009/019 Subtidal Reef Ecological Monitoring: 1999 - 2008 51 Figure 383838 Scatter plot showing the change in the proportion of trapped fine sediment (<63 µm) over time (averaged for each bay). Note: measurement of fine sediment started later in LBMMP, so there is a different time scale to Figure 34. Manly Stanmore Waiwera 80 60 40 20 Campbells Torbay Long Bay 80 Average percentage fines (<63 (<63 µm) fines percentage Average 60 40 20 Jan05 Jan06 Jan07 Jan08 Jan02 Jan03 Jan04 Jan05 Jan06 Jan07 Jan08 Jan02 Jan03 Jan05 Jan06 Jan07 Jan08 Jan02 Jan04 Jan03 Jan04 Long Bay Marine Monitoring Programme: Report 1999 – 2008 52 4.3 Examining the relationship between environmental variables and biotic abundance 4.3.1 Environmental variables and the biotic assemblage (multivariate analyses) All six environmental variables explained 46% of the variability at the bay level (Table 15). The single most important environmental variable was the fetch of each bay (‘ Fetch_bay ’), which alone explained 22.7% of the biotic variation. In the forward selection procedure, this was followed by tidally corrected depth (‘ Depth ’) which added another 12.4%, followed by average trapped sediment rate at each bay (‘ bay trap_rate’) and sediment cover (‘ sed.cover ’). These variables together explained over 43% of the biotic variation (Table 15). The rate of trapped fines (‘ bay fines_rate ’) and the short-term ENSO variable (i.e. the MEI averaged over the 3 month period prior to sampling), were also useful predictors, but these variables were dropped when a more parsimonious model was obtained using the 2nd -order corrected AIC criterion (AICc). Table 151515 Results of DISTLM analyses relating count biotic data, based on Bray-Curtis dissimilarities of log(x+1) transformed data, with environmental variables (as described in section 3.5.1). Prop. = proportion of explained variation, Cumul. = cumulative proportion of explained variation. Marginal tests Sequential tests Variable F p-value Prop. F p-value Prop. Cumul. Fetch_bay 11.75 0.001 0.227 10.23 0.001 0.227 0.227 Depth 5.78 0.001 0.126 7.4 0.001 0.124 0.350 bay trap_rate 2.54 0.013 0.060 2.8 0.006 0.046 0.395 Sed.cover 4.23 0.002 0.0954 7.4 0.013 0.037 0.432 bay fines_rate 4.46 0.001 0.126 ENSO_st 2.47 0.02 0.05 The first two axes of a dbRDA fit to the environmental variables (described in Table 15) allowed most of the fitted variation (89%) to be visualized (Figure 39).